Chimeric antigen receptor T‐cell therapy after allogeneic stem cell transplant for relapsed/refractory large B‐cell lymphoma

Abstract

Chimeric antigen receptor T-cell (CAR-T) therapy has emerged as a rapidly-advancing treatment modality for relapsed/refractory (R/R) large B-cell lymphoma (LBCL) with response rates approaching 60–80%.1, 2 The registrational clinical trials that lead to the approval of CAR-T in R/R LBCL, as well as ongoing trials, have excluded patients with previous allogeneic haematopoietic cell transplantation (alloHCT) from trial participation.3-8 To date, limited clinical data exists describing the safety and efficacy of standard-of-care CAR-T therapies following alloHCT. There are a number of potential concerns with CAR-T infusion in the context of previous alloHCT; namely, CAR-T manufacturing failure, risk of high-grade graft-versus-host disease (GVHD), and potential for increased risk of CAR-T toxicities, e.g. cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS). Here, we present our experience with post-alloHCT CAR-T with axicabtagene ciloleucel (Axi-cel) in seven patients with R/R LBCL that were heavily pretreated (see Fig 1A for specific regimens and clinical timeline). Of note, case 4 received tisagenlecleucel (Tisa-cel) prior to alloHCT on a clinical trial. All patients underwent lymphodepletion with cyclophosphamide and fludarabine before Axi-cel infusion. None of the products administered were out of specification per the United States Food and Drug Administration (FDA) label. All patients were off immunosuppression at the time of Axi-cel infusion with no evidence of GVHD. CRS was graded by the 2014 Lee criteria and ICANS was graded by Common Terminology Criteria for Adverse Events (CTCAE), version 4.9 Institutional Review Board approval was obtained for this retrospective study. Patient age ranged from 28 to 70 years, with LBCL Stages II–IV, and transplant stem cell graft donors included one matched unrelated (MUD), two matched related (MRD), and four haploidentical (Table I). Cases 1, 3, and 5 experienced Stage I–II acute GVHD peri-transplant, all managed conservatively and resolved prior to CAR-T. Donor chimerism was near 100% donor in all cases prior to CAR-T apheresis. Cases 4 and 5 each had two donor lymphocyte infusions (DLIs) months prior to Axi-cel infusion. The time elapsed between alloHCT and CAR-T therapy ranged from 5 months to 13 years. Only case 3 experienced high-grade (>2) CRS or ICANS. Following CAR-T infusion, cases 1, 2, and 6 experienced acute GVHD, with cases 1 and 6 being managed conservatively (psoralen and ultraviolet A, and oral steroids) and case 2 being severe, although not biopsy confirmed and in the setting of checkpoint inhibitor therapy (Fig 1b,c). Following CAR-T, all evaluable cases had 100% donor chimerism (unfractionated). Four patients (cases 1, 4, 5, 7) achieved a complete remission (CR), one (case 3) had a partial response (PR), and two (cases 2 and 6) had progressive disease (PD). Five of the seven cases had cytopenias at day +30, which persisted in two at day +90. At last follow-up, cases 1, 5, and 7 had no evidence of disease (NED), cases 2, 3, and 4 died due to PD, and case 6 was undergoing salvage therapy for PD. GVHD Pre-CAR-T Chimerism Pre-CAR-T (unfractionated) As our centre’s experience with CAR-T has matured, with now >100 treated patients, we have had seven patients with R/R LBCL with safe administration of CAR-T products following alloHCT. Manufacturing failure occurred in one instance (case 2) due to the product being out of specifications for release, and manufacturing was successful on the second attempt. In our experience, the use of CAR-T in patients with R/R LBCL following alloHCT has been safe and without significantly increased incidence of high-grade GVHD. Three of seven cases developed GVHD following CAR-T; however, one was in the setting of programmed cell death protein 1 (PD-1) monoclonal antibody therapy and the other two were confined to cutaneous GVHD that responded well to therapy. All three cases experiencing post-CAR-T GVHD had a time elapsed from alloHCT to CAR-T infusion of <10 months, whereas the four cases not experiencing GVHD had >12 months of elapsed time. Interestingly, the two cases that had DLI prior to CAR-T did not develop GVHD. Kochenderfer et al.10 and Park et al.11 previously reported no incidence of GVHD in CAR-T following alloHCT, from donor-derived CAR-T and acute lymphoblastic leukaemia, respectively. Thus, while the risk of GVHD following CAR-T in our experience has been greater than previously reported, we believe that CAR-T therapy following alloHCT remains safe, particularly when given >12 months after alloHCT. Furthermore, the incidence of GVHD observed in our present series is comparable to that previously reported for patients with LBCL undergoing alloHCT.12, 13 Although our present sample size is small, we did not observe a difference in CAR-T response or incidence of GVHD between MRD, MUD, and haploidentical alloHCT. In regard to toxicity, there were no treatment-related deaths and the incidence and severity of CRS and ICANS with CAR-T following alloHCT does not appear to be increased. Only case 3 experienced CRS and ICANS greater than grade 2 and this was in the setting of bulky disease with a high tumour burden. Cytopenias were present at day +30 in five of the seven patients, but persisted in only two patients at day +90 and none experienced clinically significant infections following CAR-T. Although by their very nature, patients with LBCL undergoing alloHCT have extremely aggressive and refractory disease, four of seven achieved a best response of CR. Interestingly, case 4 had CAR-T both prior to and after alloHCT (Tisa-cel clinical trial and commercially available Axi-cel, respectively), and achieved a better response (CR at day +30) with the post-alloHCT CAR-T infusion (CD19 expression was confirmed by immunohistochemistry on biopsy following relapse). However, this raises an important question of whether CAR-T cells manufactured from donor-derived engrafted lymphocytes may have greater efficacy given the possibility of graft-versus-lymphoma effect.14 Brudno et al.15 demonstrated the safe use of donor-derived T cells for manufacturing of true allogeneic CAR-T cells. However, obtaining mononuclear cells from the patient with high donor chimerism offers logistical advantage over truly donor-derived T cells, especially in the case of an unrelated donor. In summary, our present case series demonstrates that CAR-T therapy can be a safe and well-tolerated treatment modality in R/R LBCL following alloHCT, although ultimately, its efficacy compared to pure autologous CAR-T requires further study. Forat Lutfi, Jonathan Siglin, Ali Bukhari, and Noa Holtzman wrote the manuscript and constructed the tables/figures. Moaath Mustafa Ali, Dong Won Kim, Gabriela Sanchez-Petitto, and David Gottlieb assisted with reviewing and compiling medical records of patients involved, as well as manuscript editing. Kathleen Ruehle, Eilzabeth Hutnick, Natalie Gahres, and Kim Hankey were responsible for identifying the patients in our study and maintaining the CAR-T IRB-approved database at our medical centre. Seung T. Lee, Mehmet H. Kocoglu, Jean A. Yared, and Nancy M. Hardy were involved with the writing, editing, and review of this manuscript. Aaron P. Rapoport and Saurabh Dahiya were the primary individuals behind the idea for this manuscript and involved in its writing and editing. All authors had key roles in the creation of this manuscript, as noted above, and this is reflected in the order of authorship. There are no conflicts of interest by any authors involved in this manuscript.

Topics & Concepts